In 1988, 100 12- to 13-year-old children from two neighborhoods in Montevideo with different standards of living were examined. Twenty-three children with large numbers of mutans streptococci, lactobacilli, or decayed surfaces or a combination of these factors were considered to be at high risk of developing caries. Twelve of these 23 children were selected for special preventive measures. A second examination was made 18 months later, at which 81 of the originally 100 children were studied. The clinical and microbiologic variables were tested for their ability to predict caries by calculating their sensitivity, specificity, the positive and negative predictive values. The incidence of new DMFS was significantly higher in the children from the low than in those from the high socioeconomic area. It was also significantly higher in the children considered being at high risk than in those at low risk at the start of the study. Preventive measures reduced the caries risk. The highest sensitivity was obtained with the clinical test, whereas the highest specificity was obtained with the combined clinical and microbiologic tests. The findings indicate that methods for identifying children at risk which started to be used in. Scandinavia 20 years ago could be applied in the Uruguayan population of today. It is valuable to know this, as the resources for both treatment and prevention are very limited.